The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill directs the agent to install the Membrane CLI tool (@membranehq/cli) from the official NPM registry. This package is maintained by the skill's author and is required for interacting with the Membrane platform.
[COMMAND_EXECUTION]: The instructions rely on executing various membrane CLI commands to perform authentication (membrane login), manage database connections (membrane connection ensure), and interact with the EdgeDB API (membrane action run, membrane request).
[DATA_EXFILTRATION]: The skill facilitates the transfer of data between the user's environment and EdgeDB through Membrane's proxy infrastructure. This process involves network communication with getmembrane.com and edgedb.com to manage records and execute queries, which is the primary intended functionality.
[PROMPT_INJECTION]: The skill ingests data from EdgeDB query results and API responses. While this presents an indirect prompt injection surface if the database contains adversarial content, the skill uses these outputs for legitimate integration tasks without bypassing safety guidelines. (Evidence: Ingestion points: membrane action run and ExecuteEdgeQL output; Boundary markers: Absent; Capability inventory: CLI command execution and network requests; Sanitization: Absent).

edgedb